CONFIDENTIAL UNTIL PUBLISHED
FAST TRACK APPRAISAL: COST COMPARISON CASE
Aflibercept for treating myopic choroidal neovascularisation
Produced by Aberdeen HTA Group
Authors Miriam Brazzelli1
Ewen Cummins3
Shona Fielding2
Moira Cruickshank1
Cynthia Fraser1
Augusto Azuara-Blanco4
1 Health Services Research Unit, University of Aberdeen, UK
2 Medical Statistics Team, University of Aberdeen, UK
3 McMDC Ltd. Health Economics, Glasgow, UK
4 Centre for Public Health, Queen’s University Belfast
Correspondence to Miriam Brazzelli
Senior Research Fellow
University of Aberdeen
Health Services Research Unit
2nd Floor, Health Sciences Building,
Foresterhill, Aberdeen, UK, AB25 2ZD
Email: [email protected]
Date completed 18 July 2017
Version 1
Contains AIC/CIC information
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
ii
Source of funding: This report was commissioned by the NIHR HTA Programme as
project number 16/54/09.
Declared competing interests of the authors
None.
Acknowledgements
The authors are grateful to Lara Kemp for her secretarial support.
Rider on responsibility for report
The views expressed in this report are those of the authors and not necessarily those of
the NIHR HTA Programme. Any errors are the responsibility of the authors.
This report should be referenced as follows:
Brazzelli M, Cummins E, Fielding S, Cruickshank M, Fraser C, Azuara-Blanco A.
Aflibercept for treating myopic choroidal neovascularisation. Aberdeen HTA Group,
2017.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
iii
Table of contents
List of tables vi
List of figures vi
List of abbreviations vii
1 Evidence of comparable health benefits and safety 1
1.1 Is the technology pharmacologically similar to the
comparator(s)?
1
1.2 Does the technology have a marketing authorisation in the
UK? And is the marketing authorisation the same as the
NICE-recommended comparator(s)?
2
1.2.1 Is the company’s decision problem consistent with the scope? 3
1.2.2 Does the company’s decision problem cover all or only part of
the technology’s marketing authorisation for this indication?
7
1.2.3 Does the company’s decision problem cover all or only part of
the population for whom the comparator has been
recommended by NICE?
7
1.3 Has the company made a comparison to a relevant NICE-
recommended comparator?
7
1.4 Has the company positioned the technology as expected in the
treatment pathway and is the population(s), and any
subpopulation(s), defined as expected?
8
1.5 Is the claim for clinical similarity supported through the
scoping consultation?
9
1.6 Are the outcome measures and definitions of the trials for the
treatment the same to those for the NICE-recommended
comparator(s)? Or are they at least similar/comparable?
10
1.7 Strength of the clinical evidence provided by the company for
clinical similarity
17
1.7.1 Summary of evidence of aflibercept 17
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
iv
1.7.2 Critique of evidence for aflibercept 18
1.7.3 Summary of evidence against its comparators using an indirect
treatment comparison
18
1.7.4 Critique of the indirect treatment comparison 20
1.8 Is the claim for the toxicity/adverse event profile of the
technology similar to the NICE-recommended comparators?
22
1.8.1 Safety profile of aflibercept from MYRROR 22
1.8.2 Safety profile of aflibercept compared with ranibizumab 24
1.9 Is the treatment likely to offer similar or improved health
benefits over the NICE recommended comparator(s)
24
2 Evidence to support the cost-comparison case 25
2.1 Is the acquisition cost of the technology, for 1 course of
treatment, similar to/lower than the comparator
25
2.2 Are the healthcare resource costs associated with the
technology likely to be similar to/lower than the respective
costs in the NICE recommended comparator(s) appraisal
25
2.3 Has the company used the same data sources for resource
costs as the NICE recommended comparator(s)? If so, do
these reflect the most up-to-date information available from
these sources?
25
2.4 Are consequences of incorrect decision low? 25
2.5 Has the original literature search been updated? 26
3 ERG’s summary of company’s cost comparison case 27
3.1 Multiple treatments from a single vial or syringe 27
3.2 Injection frequency 27
3.3 Injection frequency and patient drop outs 30
3.4 Administration costs 32
3.5 Monitoring frequency 33
3.6 Bilateralism and recurrence 34
3.7 Cost comparison results 34
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
v
3.8 Differentiating aflibercept from ranibizumab in a cost utility
analysis
38
4 ERG’s overall view of the company’s cost
comparison case
41
5 References 42
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
vi
List of tables
Table 1 Comparison of NICE final scope and decision problem
addressed by the company
4
Table 2 Study characteristics of MYRROR and RADIANCE 11
Table 3 Outcome estimates from MYRROR (full analysis set,
LOCF)
18
Table 4 ITC results for mean 3-month gain in BCVA 19
Table 5 Proportion of participants with letter gain and loss at 3
months in the three trials
22
Table 6 Overall adverse event profile through week 24 and week
48 (reproduced from Table 24 of Document B of the
company’s submission)
23
Table 7 Previous aflibercept submissions: 1st year dosing
assumptions
27
Table 8 Company resource use survey: injection frequency
medians
28
Table 9 Company trial injection estimates, standard deviations
and 95% CIs
29
Table 10 MYRROR and RADIANCE patient flow 30
Table 11 MYRROR and RADIANCE dosing 31
Table 12 Company model per patient: at ranibizumab and
aflibercept list prices
35
Table 13 ERG sensitivity analyses: at ranibizumab and aflibercept
list prices
36
Table 14 Company indirect treatment comparison - mean letters
gain in BCVA treated eye at 3 months
38
List of figures
Figure 1 Company’s proposed treatment pathway in mCNV,
adapted from Wong et al 2015
9
Figure 2 ITC results for mean 3-month gain in BCVA 20
Figure 3 1st year mean numbers of injections 29
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
vii
List of abbreviations
AE Adverse event
AMD Age-related macular degeneration
ATA Abbreviated technology appraisal
ATE Arterial thromboembolic event
BCVA Best corrected visual acuity
BRVO Branch retinal vein occlusion
CI Confidence interval
CHMP Committee for Medicine Products for Human Use
CRT Central retinal thickness
CRVO Central retinal vein occlusion
CNV Choroidal neovascularisation
D Dioptres
DME Diabetic macular edema
DMO Diabetic macular oedema
DNA Deoxyribonucleic acid
EMA European Medicines Agency
EOT End of trial
EQ5D Euroqol 5 dimensions
ERG Evidence review group
ETDRS Early treatment diabetic retinopathy study
FA Fluorescein angiography
FAD Final appraisal determination
HRQoL Health-related quality of life
IOP Intra ocular pressure
ITC Indirect treatment comparison
IVT Intravitreal
LOCF Last observation carried forward
LTFU Long-term follow-up
mCNV Myopic choroidal neovascularisation
MO Macular oedema
NEI VFQ-25 National eye institute visual functioning questionnaire - 25
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
viii
NHS UK National Health Service
NICE National Institute for Health and Care Excellence
NMA Network meta-analysis
OCT Ocular coherence tomography
PDT Photodynamic therapy
PlGF Placental growth factor
PM Pathological myopia
PRN Pro re nata
QALY Quality adjusted life year
RCT Randomised controlled trial
RVO Retinal vein occlusion
SD Standard deviation
SmPC Summary of product characteristics
STA Single technology appraisal
TEAE Treatment-emergent adverse event
VA Visual acuity
vPDT Verteporfin photodynamic therapy
VEGF Vascular endothelial growth factor
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
1
1 Evidence of comparable health benefits and safety
1.1 Is the technology pharmacologically similar to the comparator(s)?
Vascular endothelial growth factor (VEGF) is a potent, endothelial cell mitogen that
stimulates proliferation, migration and tube formation, thus promoting angiogenic
growth of new blood vessels.1-3 VEGF has been shown to play an important role in
the development and progression of neovascularization in the eye, including in eyes
with myopic choroidal neovascularisation (mCNV).4-6 Eyes with active mCNV have
higher levels of VEGF in the aqueous humour than control eyes.7 Aflibercept and
ranibizumab are both anti-VEGF therapies.
Aflibercept (Eylea®, Bayer plc, Berkshire, UK) is a recombinant fusion protein
formed by fusing portions of human VEGF receptor 1 and 2 extracellular domains and
the Fc portion of human IgG1. It has a longer half-life in the eye than ranibizumab or
bevacizumab and a higher binding affinity to VEGF-A, as well as other VEGF
variants, including VEGF-B and placental growth factors (PlGF) 1 and 2.8-12
Aflibercept acts as a soluble decoy receptor that binds VEGF-A and PlGF and, thus,
can inhibit the binding and activation of these related VEGF receptors.10, 12, 13
Aflibercept is the only available anti-VEGF drug that acts against PlGF.12 Adverse
events classified as very common (i.e. ≥1/10) associated with aflibercept include
reduced visual acuity, conjunctival haemorrhage and eye pain. Common adverse
events (i.e. ≥1/100 to < 1/10) include detachment of the retinal pigment epithelium,
retinal degeneration, vitreous haemorrhage and cataract.
Ranibizumab (Lucentis®, Novartis Europharm Ltd, Camberley, UK) is a humanised
monoclonal antibody fragment produced in Escherichia coli cells by recombinant
DNA technology.14 Ranibizumab is a high affinity recombinant antigen that
neutralises all isoforms of VEGF-A. Binding of VEGF-A to its receptors leads to
endothelial cell proliferation and neovascularisation, as well as vascular leakage, all of
which are thought to contribute to the pathophysiology of mCNV. As ranibizumab
binds with high affinity to the VEGF-A isoforms, it prevents binding of VEGF-A to
its receptors.15 Very common eye-related adverse reactions to ranibizumab include
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
2
vitritis, vitreous detachment, retinal haemorrhage, visual disturbance, eye pain,
vitreous floaters and conjunctival haemorrhage.
The Committee for TA298 considered that anti-VEGF treatments were a substantial
improvement over previous treatments and that this improvement applied to the class
of drugs.16
Verteporfin photodynamic therapy (vPDT) was specified in the NICE final scope as a
comparator but was not included in the company’s submission. Verteporfin
(Visudyne®, Novartis Europharm Ltd, Camberley, UK) is not pharmacologically
similar to aflibercept; it is a photosensitising drug which is injected intravenously and
activated focally by illumination with light from a laser source at a wavelength
corresponding to an absorption peak of the drug. This causes a photochemical reaction
which results in direct cellular injury to vascular endothelial cells and subsequent
vessel thrombosis, thereby inducing occlusion of the CNV.17 Evidence from the VIP
trial showed that vPDT stabilised visual acuity but did not improve it at the 24-month
follow-up.18, 19 Adverse events associated with vPDT include visual disturbance,
injection site events, allergic reactions, photosensitivity reactions and chorioretinal
atrophy.18-21
1.2 Does the technology have a marketing authorisation in the UK? And is
the marketing authorisation the same as the NICE-recommended
comparator(s)?
Aflibercept has UK marketing authorisation (since 28th October 2015) for adults for
the treatment of neovascular (wet) age-related macular degeneration (AMD), visual
impairment due to macular oedema secondary to RVO (BRVO or CRVO), visual
impairment due to diabetic macular oedema (DMO) and visual impairment due to
myopic choroidal neovascularisation (CNV).
Ranibizumab has UK marketing authorisation for adults for the treatment of
neovascular (wet) AMD, visual impairment due to DMO and visual impairment due
to macular oedema secondary to retinal vein occlusion (BRVO or CRVO), visual
impairment due to choroidal neovascularisation (CNV). Ranibizumab was approved
by NICE for this indication on 27th November 2013 (TA298).16
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
3
The Eylea® summary of product characteristics (SmPC) specifies “myopic choroidal
neovascularisation”,22 and the Lucentis® SmPC states “choroidal
neovascularisation”.14
Verteporfin photodynamic therapy has UK marketing authorisation for the treatment
of adults with exudative (wet) age-related macular degeneration with predominantly
classic subfoveal CNV or adults with subfoveal CNV secondary to pathological
myopia.
The company’s decision problem was not consistent with the comparators specified in
the final scope issued by NICE, in that the company included ranibizumab as a
comparator but did not include vPDT. The company’s justification, that vPDT is not
standard treatment in the NHS for mCNV, was previously noted by the ERG involved
in TA29816 and is further acknowledged by the ERG in this assessment (the same
ERG in both cases). In addition, vPDT is indicated for subfoveal mCNV only and,
therefore, suitable for only part of the population of adults with visual impairment due
to mCNV which is specified in the scope (albeit the proportion of patients with
subfoveal mCNV – as compared to juxtafoveal or extrafoveal mCNV – tends to be at
least half of the overall population with mCNV.23
1.2.1 Is the company’s decision problem consistent with the scope?
Table 1 presents a comparison of the NICE final scope and the decision problem
addressed by the company.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
4
Table 1 Comparison of NICE final scope and decision problem addressed by the company
Final scope issued by NICE Decision problem addressed in the submission
Comments from the company
Comments from the ERG
Population Adults with visual impairment due to myopic choroidal neovascularisation
Adults with visual impairment due to myopic choroidal neovascularisation
None None
Intervention Aflibercept Aflibercept None None
Comparators • Ranibizumab • Verteporfin photodynamic
therapy
• Ranibizumab Bayer considers that the most appropriate comparator is ranibizumab (Lucentis). Ranibizumab (an alternative anti-VEGF therapy) has been appraised by NICE in this indication (TA 298)16 Verteporfin photodynamic therapy (vPDT) is not an appropriate comparator as it is not standard treatment within the NHS for mCNV. It was acknowledged by the Evidence Review Group (ERG) during appraisal of ranibizumab in this indication, that vPDT was rarely used in clinical practice.
The ERG agrees that ranibizumab is the most appropriate comparator and that vPDT is not an appropriate comparator as it is rarely used in UK clinical practice for mCNV
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
5
Final scope issued by NICE Decision problem addressed in the submission
Comments from the company
Comments from the ERG
Outcomes • Best corrected visual acuity (affected eye)
• Best corrected visual acuity • (both eyes) • Contrast sensitivity • Adverse effects of treatment • Health-related quality of life
• Best corrected visual acuity (affected eye)
• Adverse effects of treatment • Health-related quality of life In addition to the outcomes proposed in the scope, Bayer presented on: • Proportion of patients gaining
≥15 ETDRS letters at week 24 from baseline
• Mean change from baseline in BCVA score at each visit and at week 48
• Proportion of patients gaining or losing ≥15, ≥10 or ≥5 ETDRS letters at week 48 from baseline
• Ad-hoc analysis of exposure (as relevant to the indirect comparison with ranibizumab).
Bayer will not be presenting data on contrast sensitivity, as listed in the pre-invitation scope, as this was not collected in the pivotal study. Bayer will also not be presenting data on best corrected visual acuity (both eyes) as in the pivotal study, only one eye was designated as the study eye and BCVA (both eyes) was not assessed.
Further outcome data are presented to further report on the efficacy of aflibercept. Exposure is presented as it is relevant to the indirect comparison with ranibizumab.
The ERG agrees with the company’s justification for not presenting data on contrast sensitivity at it is not used in clinical practice to make decisions. In the case of BCVA (both eyes), the ERG considers that it would be useful to have scores for both eyes, but that the affected eye is sufficient information for the purposes of testing equivalence
Economic analysis
The reference case stipulates that the cost effectiveness of treatments should be expressed in terms of incremental cost per quality-adjusted life year.
In light of the consultation on the ATA process for appraisal, Bayer considers that the most appropriate economic evaluation should be based on a cost-
As discussed at the decision problem meeting
The ERG considers the company’s approach to be justified because of similar efficacy and safety.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
6
Final scope issued by NICE Decision problem addressed in the submission
Comments from the company
Comments from the ERG
The reference case stipulates that the time horizon for estimating clinical and cost effectiveness should be sufficiently long to reflect any differences in costs or outcomes between the technologies being compared. Costs will be considered from an NHS and Personal Social Services perspective. The availability of any patient access schemes for the intervention or comparator technologies will be taken into account. Cost effectiveness analysis should include consideration of the benefit in the best and worst seeing eye.
comparison analysis compared to standard of care ranibizumab
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
7
1.2.2 Does the company’s decision problem cover all or only part of the
technology’s marketing authorisation for this indication?
The company’s decision problem covers all of the marketing authorisation for this
indication.
1.2.3 Does the company’s decision problem cover all or only part of the
population for whom the comparator has been recommended by NICE?
NICE TA29816 states that “ranibizumab is recommended as an option for treating
choroidal neovascularisation associated with pathological myopia when the
manufacturer provides ranibizumab with the discount agreed in the patient access
scheme”. The Lucentis® summary of product characteristics states that it is indicated
for adults and states that “The safety and efficacy of Lucentis in children and
adolescents below 18 years of age have not been established”.14 However, it is worth
noting that myopic choroidal neovascularization is extremely rare in children/young
people under 18 years of age. Verteporfin PDT was specified in the NICE final scope
as a comparator. Visudyne® is indicated for “adults with subfoveal choroidal
neovascularisation secondary to pathological myopia”. Thus, this comparator is only
indicated for part of the population for whom aflibercept is indicated.
1.3 Has the company made a comparison to a relevant NICE-recommended
comparator?
Yes. Ranibizumab was recommended by NICE on 27th November 2013 “as an option
for treating visual impairment due to CNV secondary to pathological myopia when
the manufacturer provides ranibizumab with the discount agreed in the patient access
scheme”16 Ranibizumab is the only anti-VEGF therapy currently approved by NICE
for this indication. Bevacizumab is another anti-VEGF therapy that was used off-
license in clinical practice prior to the approval of ranibizumab but is not licensed for
any eye conditions. Its use has now all but ceased following the approval of
ranibizumab for this indication. Verteporfin PDT is appropriately licensed for this
condition (subfoveal CNV only) and was widely used in clinical practice in the past.
However, it has now been superseded by anti-VEGF therapies which have showed
superior gains in visual acuity24 without the development of chorioretinal atrophy
associated with vPDT.20, 21
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
8
1.4 Has the company positioned the technology as expected in the treatment
pathway and is the population(s), and any subpopulation(s), defined as expected?
Yes. The company has positioned aflibercept as expected in the proposed treatment
pathway. There are currently no definitive guidelines for treating mCNV. The
company’s submission reproduced a treatment algorithm for diagnosing and treating
mCNV, proposed by Wong 2014,23 and based on a summary of the current treatment
options for mCNV (i.e. laser photocoagulation, vPDT, ranibizumab, bevacizumab,
aflibercept). The ERG agrees that the treatment algorithm is an accurate reflection of
current clinical practice in the NHS. The company’s use of an immediate anti-VEGF
injection as first-line therapy is consistent with the proposed algorithm and current
UK practice; in the NHS, this can take place on the same day as the diagnosis or
within a few weeks. The ERG agrees with the company’s follow-up monitoring
strategy; the company’s version of the algorithm does not specify time points for
monitoring and the monitoring of disease activities strategy proposed by Wong
et al states: monthly for months 1 and 2 and then at least three-monthly in the
first year. The ERG clinical expert noted that the stage of disease at which
patients present is not known in NHS clinical practice. It is, therefore, unclear if
patients in the NHS are assessed at a similar stage as those recruited from
different countries in the trials.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
9
Figure 1 Company’s proposed treatment pathway in mCNV, adapted from
Wong et al 2015 (reproduced from Figure 1 of Document B of the company’s
submission)
The company’s definition of the population is as would be reasonably expected, based
on the marketing authorisation of aflibercept.
1.5 Is the claim for clinical similarity supported through the scoping
consultation? (Refer back to the scope and decision problem pro-forma)
The company claimed clinical similarity during the decision problem phase. In their
decision problem pro-forma they stated:
“Bayer considers that the appraisal of aflibercept for treating myopic choroidal
neovascularisation could be conducted under the proposed Abbreviated Technology
Appraisal (ATA) process. Whilst the process for conducting such an appraisal has not
been finalised and published, Bayer would be happy to proceed on this basis, subject
to NICE agreement. Aflibercept in this indication would appear to meet the criteria
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
10
for such an appraisal – ‘We propose to use this process for new technologies that
provide similar or greater health benefits, compared with existing NICE-
recommended technologies at a similar or lower cost’
https://www.nice.org.uk/about/what-we-do/our-programmes/nice-guidance/nice-
technology-appraisal-guidance/abbreviated-technology-appraisal-process-
consultation”.
1.6 Are the outcome measures and definitions of the trials for the treatment
the same to those for the NICE-recommended comparator(s)? Or are they at
least similar/comparable?
The outcomes specified in the NICE final scope and considered by the company were
consistent with the outcomes considered in RADIANCE/TA298 (i.e. BCVA for the
affected eye, adverse effects of treatment, HRQoL).16, 25 In addition, the RADIANCE
trial considered the proportion of participants with gain or loss of 5, 10 and 15 letters,
which proved to be a major driver of the economic model. The company provided the
relevant data to the ERG at clarification. The scopes for both the present appraisal and
TA298 also specified BCVA (both eyes) and contrast sensitivity as outcomes but
neither submission considered them. The ERG agrees that it was appropriate for
the company not to consider these outcomes, as (i) contrast sensitivity is not used
in clinical practice to make decisions and (ii) whilst it would be clinically useful
to have information about BCVA in both eyes, the affected eye is probably
sufficient for the purposes of testing equivalence.
RADIANCE has two ranibizumab arms, one with retreatment based on disease
activity and one based on visual acuity stabilisation.25 Details are presented in Table
2. The company stated in its clarification response that neither ranibizumab arm is
particularly similar to the aflibercept arm of MYRROR in terms of retreatment
criteria. The ERG clinical expert is of the opinion that the disease activity arm in
RADIANCE - where vision impairment is one of the criteria of disease activity -
is most comparable with the aflibercept arm in MYRROR.26
The company conducted risk of bias assessments of the MYRROR and RADIANCE 25, 26 trials based on the Cochrane risk of bias domains. The ERG is satisfied that
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
11
randomisation and masking were adequate in both trials and that they were, in
general, at low risk of bias.
Table 2 Study characteristics of MYRROR and RADIANCE
MYRROR26 RADIANCE25
Study design Phase III, multicentre, randomised,
double-masked, sham-controlled
study
Phase III, multicentre,
randomised, double-masked,
active-controlled study
Location/no of
centres
20 centres in Asia (Hong Kong,
Japan, Republic of Korea,
Singapore, Taiwan)
76 centres worldwide (Austria,
Canada, France, Germany,
Hong Kong, Hungary, India,
Italy, Japan, Latvia, Lithuania,
Poland, Portugal, Singapore,
Slovakia, South Korea, Spain,
Switzerland, Turkey, UK)
Intervention Intravitreal aflibercept 2.0mg
Day 1-week 20: IVT aflibercept at
baseline, then PRN dosing every 4
weeks of IVT aflibercept 2mg or
sham injection, subject to meeting at
least one of the following re-
treatment criteria:
(i) Reduction in VA by ≥5
letters from the previous
ETDRS examination;
(ii) Increase in CRT >50µm
from the time of the
previous examination,
new or persistent cystic
retinal changes,
subretinal fluid, or
pigment epithelial
detachment, and new or
persistent CNV or
bleeding; or
Intravitreal ranibizumab
0.5mg (guided by disease
activity criteria or VA
stabilisation criteria)
Day 1: Both ranibizumab
groups received IVT
ranibizumab 0.5mg
Disease activity group: from
month 1 onwards, dosing was
stopped if no disease activity
was seen (i.e. vision
impairment, attributable to intra
or subretinal fluid or active
leakage secondary to PM as
assessed by OCT and/or FA).
Treatment was resumed when
the disease activity criterion
was fulfilled and continued until
no disease activity was seen
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
12
MYRROR26 RADIANCE25
(iii) Deemed necessary by
the investigator based on
their clinical impression
or diagnostics
performed in the context
of standard medical
care.
Week 24-week 44: PRN dosing of
IVT aflibercept 2mg or sham
injection, subject to meeting at least
one of the above retreatment criteria
Stabilisation group: IVT
ranibizumab 0.5mg at month 1.
For the following months,
treatment was stopped if the
stabilisation criterion for BCVA
was fulfilled (i.e. no change in
BCVA as compared with the
two preceding monthly visits).
Treatment was resumed with
monthly injections when there
was a loss of BCVA due to
disease activity and was
continued until stable BCVA
was re-established for three
consecutive monthly
assessments
Comparator Sham injections
Day 1-week 20: sham injection at
baseline, then sham injections every
4 weeks
Week 24-week 44: mandatory IVT
aflibercept 2mg, then PRN dosing
every 4 weeks of IVT aflibercept
2mg or sham injection, subject to
meeting at least one of the above
retreatment criteria
vPDT
Day 1: 6mg/m2 intravenously
followed by a standard fluence
rate of 600 mW/cm2 delivered
for 83 seconds with a light dose
of 50 J/cm2
Months 3-11: patients with
disease activity could receive
ranibizumab 0.5mg, vPDT, or
ranibizumab 0.5mg plus vPDT.
Treatment was stopped if no
disease activity was seen.
Treatment was resumed when
the above disease activity
criterion was fulfilled and
continued until no disease
activity was seen
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
13
MYRROR26 RADIANCE25
No of
participants
randomised
Aflibercept 2.0mg: n=91
Sham: n=31
Total: n=122
Ranibizumab 0.5mg (disease
activity criteria): n=116
Ranibizumab 0.5mg
(stabilisation criteria): n=106
vPDT: n=55
Total: n=277
Main inclusion
criteria
≥18 years of age
High myopia, defined as ≤-6.0
dioptres or axial length of ≥26.5mm
Active (defined by leakage on
fluorescein angiography) subfoveal
or juxtafoveal (within 1-199µm from
the centre of the fovea) myopic
CNV
BCVA of 73-35 letters (ETDRS
equivalent of 20/40-20/200) at 4m
≥18 years of age
Diagnosis of active CNV
secondary to PM in the study
eye using the following criteria:
• Presence of myopia greater
than −6 D of spherical
equivalence
• Ocular ultrasonography or
biometry demonstrating
anterio-posterior elongation
measurement ≥ 26 mm
• Presence of posterior
changes compatible with the
PM seen by fundus
ophthalmoscopy and fundus
photography
• Presence of active leakage
from CNV seen by FA
• Presence of intra- or
subretinal fluid seen or
increase of CRT by OCT
At least one of the following
lesion types present in the study
eye:
• Subfoveal (presence of
abnormal neovasculature in
the avascular central fovea)
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
14
MYRROR26 RADIANCE25
• Juxtafoveal (presence of
abnormal neovasculature
not under the centre of the
fovea but < 200 μm from
the centre) with
involvement of the central
macular area
• Extrafoveal (presence of
abnormal neovasculature
more than 200 μm from the
centre of the fovea) with
involvement of the central
macular area
• Margin of the optic disc
(presence of abnormal
neovasculature at
peripapilar area) with
involvement of the central
macular area
BCVA of ≥ 24 letters and ≤ 78
letters tested at 4 m starting
distance using ETDRS-like
BCVA chart
Visual loss only due to the
presence of any eligible types of
CNV related to PM based on
clinical ocular findings
(described at inclusion criteria
of the study eye), FA and OCT
Main exclusion
criteria
Only 1 functional eye
Recurrent myopic CNV or aphakia
(including pseudophakic patients) in
study eye
History of (a) stroke, (b) pan-
retinal or focal/grid laser
photocoagulation with
involvement of the macular area
in the study eye at any time, (c)
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
15
MYRROR26 RADIANCE25
CNV with an origin other than PM
in study eye
Any iris neovascularisation or
vitreous haemorrhage in either eye
Uncontrolled glaucoma, defined as
intraocular pressure≥25mmHg on
optimal medical regimen
Previous filtration surgery in either
eye
Pregnant or nursing women
intraocular treatment with
corticosteroids or intraocular
surgery in past 3 months & anti-
VEGF or vPDT treatment at
any time in study eye, or (d)
hypersensitivity to ranibizumab
or verteporfin or drugs of a
similar class
Presence of CNV secondary to
any cause other than PM
Presence of active infectious
disease or intraocular
inflammation, active or
suspected periocular infection,
confirmed IOP≥25mmHg, or
iris neovascularisation either
eye at enrolment
Pregnant or nursing women
Primary
outcome
Mean change in BCVA from
baseline to week 24
Mean average change in BCVA
from baseline to month 1
through month 3 (defined as
mean difference of BCVA
versus baseline over all monthly
post-baseline assessments from
month 1 to month 3)
Other outcomes Proportion of patients gaining ≥15
letters at week 24
Absolute change or mean change
from baseline in CRT (as assessed
by OCT at week 24 and week 48)
Absolute change in CNV lesion size
from baseline (as assessed by FA at
week 24 and week 48)
Mean average change in BCVA
from baseline to month 1
through month 6
Mean change in BCVA from
baseline over time
Proportion of patients gaining
≥10 and ≥15 ETDRS letters (or
reaching 84 letters) at month 12
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
16
MYRROR26 RADIANCE25
Proportion of patients gaining ≥15
letters from baseline at week 48
Proportion of patients gaining ≥10
letters from baseline at week 24 and
week 48
Leakage from CNV (as assessed by
FA from baseline to week 24 and
week 48)
Change in EQ-5D score from
baseline to week 24 and week 48
Change in 25-item NEIVFQ 25 total
score from baseline to week 24 and
week 48
Proportion of patients losing
≥10 and ≥15 ETDRS letters at
month 12
Study duration 48 weeks 12 months
No of injections Aflibercept group:
Weeks 0-8: Median 2 (mean 2)
Weeks 12-20: Median 0 (mean 0.9)
Weeks 24-32: Median 0 (mean 0.8)
Weeks 36-44: Median 0 (mean 0.5)
Weeks 0-48: Median 3 (mean 4.2)
Sham group:
Weeks 0-48: Median 3 (mean 3)
Weeks 24-32 (3rd quarter):Median 2
(mean 1.8)
Weeks 36-44 (4th quarter): Median 1
(mean 1.2)
Disease activity group:
Total 404
Mean (SD) 3.5 (3)
Median 2
VA stabilisation group:
Total 488
Mean (SD) 4.6 (2.6)
Median 4
vPDT group:
Total 131
Mean (SD) 2.4 (2.6)
Median 2
1.7 Strength of the clinical evidence provided by the company for clinical
similarity
1.7.1 Summary of evidence of aflibercept
The MYRROR study provides the only RCT evidence for use of aflibercept in
treating mCNV in comparison to sham-controlled injections.26 MYRROR randomised
122 patients (91 to aflibercept, 31 to sham) across 20 study centres in Asia. Baseline
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
17
characteristics were balanced across treatment groups. The primary outcome was
mean change in BCVA from baseline to 24 weeks. Participants were followed until 48
weeks, to allow safety data to be collected alongside additional longer-term outcome
data.
By 24 weeks, the aflibercept group showed an increase of 12.1 letters from baseline,
while the sham group had lost two letters, resulting in a greater improvement in the
aflibercept group of 14.1 ETDRS letters (95% CI 10.8, 17.4; p-value < 0.0001)
adjusting for country and baseline BCVA. Aflibercept was significantly better in the
proportion of patients with ≥5, ≥10 and ≥15 ETDRS letters, and had fewer participants
with ≥5, ≥10, ≥15 ETDRS letter loss (Table 3).
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
18
Table 3 Outcome estimates from MYRROR (full analysis set, LOCF)26
Week 24
Aflibercept
(n =90)
Sham
(n = 31) Difference 95% CI p-value
Mean change in BCVA baseline
to 24 weeks 12.1 -2 14.1 (10.8, 17.4) <0.0001
Proportion ≥15 letters gain (%) 38.9 9.7 29.2 (14.4, 44.0) <0.001
Proportion ≥10 letters gain (%) 63.3 12.9 XXX XXX XXX
Proportion ≥5 letters gain (%) 83.3 19.4 XXX XXX XXX
Proportion ≥5 letters loss (%) 3.3 35.5 XXX XXX XXX
Proportion ≥10 letters loss (%) 0 25.8 XXX XXX XXX
Proportion ≥15 letters loss (%) 0 6.5 XXX XXX XXX
NEI VFQ-25 mean change
baseline to 24 weeks 3.14(XXX) -2.58 (XXX) 5.21 (1.25, 9.18) 0.010
EQ5D mean change baseline to
24 weeks 0.0187 0.0341 -0.0045 (-0.058, 0.049) 0.8690
1.7.2 Critique of evidence for aflibercept
MYRROR was conducted on an entirely Asian population which may cause some
concern regarding the applicability to a UK population.26 The ERG clinical opinion
is that the effect would not differ between ethnic groups. This view is shared by
the CHMP, who agreed the results of MYRROR can be extrapolated to the European
population, and the European Medicines Agency accepts the results of MYRROR as
representative for mCNV patients in Europe, regardless of ethnicity. This issue is
discussed further when describing the indirect treatment comparison.
Another issue with MYRROR is that efficacy data for aflibercept versus sham is only
available up until 24 weeks, as after that time sham patients were switched to
aflibercept. It would be better to have longer term outcome data; however, the ERG
opinion is that most of the effect will occur in the first few months. Safety data are
available up until 48 weeks.
1.7.3 Summary of evidence against its comparators using an indirect treatment
comparison
There are no head to head trials of aflibercept against its comparator ranibizumab. The
RADIANCE trial provides evidence of ranibizumab versus verteporfin photodynamic
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
19
therapy (vPDT).25 The latter is not considered a relevant comparator in this appraisal
as ranibizumab was approved by NICE and vPDT has been phased out of clinical
practice. The VIP study 18, 19compares vPDT with placebo and can be used as the link
between MYRROR and RADIANCE in an indirect treatment comparison, even
though we are not directly interested in the results of vPDT.
A network was established using the MYRROR, RADIANCE AND VIP trials.18, 19, 25,
26 RADIANCE had two ranibizumab arms versus vPDT.25 One group received
ranibizumab injection on day 1 and month 1 with retreatment based on visual acuity
stabilisation criteria. The second ranibizumab arm received injection on day 1 and,
starting in month one, retreatment was based on disease activity criteria. These two
arms are referred to ranibizumab visual acuity and ranibizumab disease activity,
respectively. In the VIP trial, patients were randomised to either vPDT or placebo.18, 19
Table 4 ITC results for mean 3-month gain in BCVA
Mean SD 95% low 95% high
vPDT vs placebo 1.05 2.29 -3.47 5.50
Ranibizumab visual acuity vs placebo 11.75 2.75 6.31 17.09
Ranibizumab disease activity vs
placebo 12.15 2.72 6.76 17.43
Aflibercept vs placebo 13.09 2.04 9.10 17.08
Aflibercept vs vPDT 12.04 3.05 6.10 18.00
Aflibercept vs ranibizumab visual
acuity 1.34 3.40 -5.35 8.00
Aflibercept vs ranibizumab disease
activity 0.94 3.38 -5.67 7.56
BCVA, Best Corrected Visual Acuity; ITC, indirect treatment comparison; SD,
standard deviation; vPDT, verteporfin photodynamic therapy.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
20
Figure 2 ITC results for mean 3-month gain in BCVA (reproduced from Figure
8 of Document B of the company’s submission)
The network considered the 3-month mean BCVA change in all treatments versus
placebo and versus aflibercept. Three months was used as, in RADIANCE,25 the
vPDT arm switched to ranibizumab after 3 months, so the only common endpoint that
could be considered was at 3 months. Within MYRROR,26 no 3-month end point data
were collected, but data were collected at 12 and 16 weeks, so a weighted average was
used (75% at 12 weeks and 25% of 16 week assessments). The results of the indirect
treatment comparison are shown in Table 4 (source Table 23, Document B) and
Figure 2 (for comparison with placebo). Both aflibercept and ranibizumab were found
to be significantly better than placebo, but not significantly different from each other.
1.7.4 Critique of the indirect treatment comparison
The trials are similar in the majority of patient demographics but the biggest concern
for heterogeneity is the ethnic distribution. Within VIP, 91% are Caucasian, while in
RADIANCE it is 57% Caucasian.25 The MYRROR study was conducted in an
entirely Asian population (Japan, Hong Kong, Singapore, South Korea and Taiwan).26
The company conducted their own analysis to test the sensitivity of effect in different
ethnic groups. Trials using aflibercept for other approved conditions were considered.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
21
Data on 1104 participants, 80% Caucasian, were obtained. The company reported that
absolute treatment differences between Asian people and white people showed similar
efficacy trends between groups, with overlapping confidence intervals (no actual data
given). The European Medicines Agency (EMA) accepted that the efficacy results of
MYRROR are representative for mCNV patient populations in Europe regardless of
ethnicity. The ERG clinical expert’s opinion is that the efficacy of aflibercept is
unlikely to be different in the various ethnic groups; thus, despite MYRROR
being conducted in an Asian population, the results are considered transferrable
to the UK population. This view is shared by the EMA.
The indirect comparison makes use of the 3 month data only, because of treatment
switching at this stage in the vPDT arm of RADIANCE.25 Within MYRROR,26 no 3
months end point data were collected, but data were collected at 12 and 16 weeks, so
a weighted average was used (75% at 12 weeks and 25% of 16 week assessments). It
could be argued that 12 weeks and 13 weeks (3 months) are not that different, so the
12 week data should have been included in the indirect treatment comparison, rather
than the weighted average. The ERG requested this at clarification and the company
provided an alternative ITC. The relative treatment benefit of aflibercept versus
placebo was almost the same at 12 weeks as the imputed 13 week value, so the results
of the ITC did not change significantly. The ERG accept the results originally
presented by the company (Table 4, Figure 2)
The indirect treatment comparison was not undertaken for the proportion of patients
who gained or lost 5, 10 and 15 ETDRS letters. The company indicated during the
clarification process that they considered the possibility of conducting an indirect
comparison for gaining and losing 15 ETDRS letters as VIP contained that
information. However, due to zero events (no events in placebo arm of VIP and no
events in ranibizumab arm of RADIANCE) and the limited network, this proves
unfeasible. The ERG agrees this indirect comparison is not possible and present
the relevant information in Table 4.
The proportions of patients gaining ≥ 15 letters and ≥ 10 letters and losing ≥ 15 letters
and ≥ 10 letters are very similar in the aflibercept arm and the ranibizumab arm of the
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
22
respective trials. The ERG is content that aflibercept and ranibizumab provide
similar outcomes with regard to gaining and losing ETDRS letters.
Table 5 Proportion of participants with letter gain and loss at 3 months in the
three trials
Proportion letter gain or loss
≥15
letters
gain
≥10
letters
gain
≥10
letters
loss
≥15
letters
loss
MYRROR26 Aflibercept (n =90) 38.9 63.3 0 0
Sham (n = 31) 9.7 12.9 25.8 6.5
RADIANCE25
Ranibizumab visual
acuity (n =105)
38.1 61.9 1.9
1.9
Ranibizumab disease
activity (n = 116) 43.1 65.5 0.9
0
vPDT (n=55) 14.5 27.3 16.4 7.4
VIP18, 19 vPDT (n =81) 2 - - 6
placebo (n = 39) 0 - - 21
1.8 Is the claim for the toxicity/adverse event profile of the technology similar
to the NICE-recommended comparators?
The ERG and the ERG clinical expert agree that there are unlikely to be major
differences in the safety profile of aflibercept and its comparator ranibizumab. Further
details are provided here.
1.8.1 Safety profile of aflibercept from MYRROR
The safety population included all patients who had received any study treatment.
Safety was monitored by recording ocular and non-ocular adverse events (AEs) at
each study visit (every 4 weeks). Treatment emergent AEs (TEAEs) were AEs which
occurred or worsened after the first administration of study drug and within 30 days
after the last study injection (active or sham). In total, across the 48 weeks, 64
(70.3%) aflibercept patients experienced at least one TEAE compared to 18 (58.1%)
of those in the sham group. No deaths were reported during the study. Table 6
presents the overall adverse event profile.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
23
Table 6 Overall adverse event profile through week 24 and week 48
(reproduced from Table 24 of Document B of the company’s submission) Through week 24 Through week 48
Aflibercept
(n=91)
n (%)
Sham
(n=31)
n (%)
Aflibercept
(n=91)
n (%)
Sham
+aflibercept
(n=31)
n (%)
Any TEAE XXX XXX 64 (70.3) 18 (58.1)
Non-ocular (systemic) 40 (44.0) 10 (32.3) 53 (58.2) 12 (38.7)
Ocular (study eye) 21 (23.1) 6 (19.4) 29 (31.9) 11 (35.5)
Any study drug-related TEAE XXX XXX 9 (9.9) 2 (6.5)
Ocular drug-related (study eye) XXX XXX 6 (6.6) 1 (3.2)
Non-ocular drug-related XXX XXX 3 (3.3) 1 (3.2)
Any injection-related TEAE
Injection-related ocular TEAE in
study eye
15 (16.5)
XXX
4 (12.9)
XXX
18 (19.8)
18 (19.8)
4 (12.9)
4 (12.9)
Any procedure-related TEAE XXX 0 12 (13.2) 0
Procedure-related ocular TEAE in
study eye
5 (5.5) 0 6 (6.6) 0
Any serious AE XXX XXX 7 (7.7) 1 (3.2)
Non-ocular (systemic) XXX XXX 4 (4.4) 0
Ocular (study eye)
Any serious TEAE
0
XXX
0
0
1 (1.1)
7 (7.7)
0
0
Drug-related serious TEAE 0 0 1 (1.1)
(ocular)
0
Any injection-related serious TEAE
(study eye)
0 0 1 (1.1) 0
Any procedure-related serious TEAE 0 0 1 (1.1) 0
Any AEs leading to discontinuation of
study drug
XXX XXX XXX XXX
Any death 0 0 0 0
Any ATE events XXX 0 1 (1.1) 0
AE=adverse event; ATE=arterial thromboembolic event; TEAE=treatment-emergent adverse event
Most reported ocular TEAEs were of mild intensity, resolved within the study period
and did not lead to interruption or permanent discontinuation of treatment. The most
frequently reported ocular TEAEs in the study eye were conjunctival haemorrhage
(11%), eye pain (7.7%) and punctate keratitis (6.6%) in the aflibercept group. In the
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
24
sham group, the TEAEs were punctate keratitis (12.9%), dry eye (6.5%) and posterior
capsule opacification (6.5%) [source: Company Submission Document B, Table 25].
Non-ocular TEAEs were reported in 44% of aflibercept patients and 32.3% sham
patients ay 24 weeks, rising to 58.2% and 38.7%, respectively by 48 weeks. The most
common non-ocular TEAEs reported by 48 weeks were nasopharyngitis (aflibercept
18.7%, sham+aflibercept 9.7%), headache (aflibercept 6.6%, sham+aflibercept 3.2%),
and nausea (aflibercept 7.7%, sham+aflibercept, 0%).
The company report that aflibercept injections are well tolerated and the safety profile
of aflibercept for mCNV was generally consistent with the known safety profile for
the other conditions for which it is licensed (wet AMD, RVO, DMO). The ERG and
the ERG’s clinical expert agree with the company that the safety profile is
comparable to that observed in other eye conditions.
1.8.2 Safety profile of aflibercept compared with ranibizumab
No head-to-head safety information is available for mCNV and the VIP trial did not
collect relevant data to allow an indirect comparison. The company summarised the
findings from VIEW 1 and VIEW 2, which compared aflibercept with ranibizumab
for wet AMD. These studies demonstrated aflibercept to be well tolerated with
comparable safety profile to ranibizumab in relation to ocular and non-ocular adverse
events. Data in the VIEW trials were available for 2 years providing a longer term
comparison. This information provides evidence that aflibercept and
ranibizumab are similar in their safety profile and we do not anticipate that to
change across eye conditions.
1.9 Is the treatment likely to offer similar or improved health benefits over
the NICE recommended comparator(s), e.g. similar/fewer adverse events,
similar/improved clinical outcomes?
The ERG’s opinion is that aflibercept is likely to offer similar benefits to ranibizumab
with a similar safety profile.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
25
2 Evidence to support the cost-comparison case
2.1 Is the acquisition cost of the technology, for 1 course of treatment, similar
to/lower than the comparator?
The list price of an aflibercept vial is £816. The list price of a ranibizumab pre-filled
syringe and vial are both £551. If vials are used, needle and syringe costs will be
minor. The aflibercept vial only contains enough for one administration. Both the
ranibizumab pre-filled syringe and vial could theoretically be used for more than one
administration which would result in significant cost savings, but ERG clinical expert
opinion is that this does not occur in the NHS.
2.2 Are the healthcare resource costs associated with the technology likely to
be similar to/lower than the respective costs in the NICE recommended
comparator(s) appraisal?
The ranibizumab pre-filled syringe is more convenient to use than the aflibercept vial.
ERG clinical expert opinion is that this does not affect the numbers of patients that
can be seen in clinic. The administration cost per injection will be the same for
aflibercept and ranibizumab as will the monitoring costs while patients are on
treatment.
2.3 Has the company used the same data sources for resource costs as the
NICE recommended comparator(s)? If so, do these reflect the most up-to-date
information available from these sources?
The company has assumed that the same number of injections will be required
whether aflibercept or ranibizumab is used. As a consequence, it does not need to
consider administration and monitoring costs.
2.4 Are consequences of incorrect decision low?
Given the results of the company NMA the clinical consequences of an incorrect
decision are minimal.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
26
Under the company assumptions and applying list prices there is a net cost from
aflibercept of £6,057 per patient due to £265 higher drug cost per administration. A
50% market share suggests an annual present value cost of £7.9mn.
Applying the mean numbers of injections from the trials for the 1st year of 4.2
injection for aflibercept and 3.5 for the ranibizumab disease activity arm and
including a net cost of £55 per additional injection suggests a net cost of £8,257 per
patient. A 50% market share suggests an annual present value cost of £10.7mn. If the
statistically significant higher drop-out for aflibercept is adjusted for, the worst case
scenario, this might suggest 4.5 aflibercept injections in the 1st year and a net cost of
£9,612 per patient. A 50% market share suggests an annual present value cost of
£12.5mn. These estimates are sensitive to the XXXX recurrence rate. It also shows
some sensitivity to the baseline XXXX bilateralism rate, which is assumed to be
constant.
Crude patient benefit calculations by the ERG, outlined in greater detail in the final
summary section below, suggest that at central clinical and resource use estimates
aflibercept is unlikely to be cost effective at list prices regardless of whether the better
seeing eye is treated or only the worse seeing eye is treated.
2.5 Has the original literature search been updated?
A sensitive search was undertaken to identify relevant studies on any treatment for
myopic choroidal neovascularisation. No date limits were applied to the main
searches and date of last search was 22nd November 2016. The company did not rely
on the literature searching that had been undertaken for the previous assessments.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
27
3 ERG’s summary of company’s cost comparison case
3.1 Multiple treatments from a single vial or syringe
Aflibercept is only available as a 0.1ml 4mg vial. The dose per administration is 2mg.
Given the need for some headroom, only one administration is possible with each
aflibercept vial.
Ranibizumab is available as both a 0.23ml 2.3mg vial and as a 0.165ml 1.65mg
prefilled syringe. The dose per administration is 0.5mg. The ERG clinical expert has
indicated that both ranibizumab vials and syringes could, in theory, be used to treat
more than a single eye. It might be possible to treat a patient bilaterally using a single
vial, or even to treat more than one patient using a single vial. This could significantly
reduce the costs of using ranibizumab compared to aflibercept.
The ERG clinical expert suggests that multiple injections of different patients from a
single ranibizumab vial or syringe is not current NHS practice. The ERG clinical
expert suggests that bilateral treatment of a single patient from a single ranibizumab
vial or syringe is not current NHS practice.
3.2 Injection frequency
The company applies the same number of injections for aflibercept and ranibizumab.
The company argues that since the mean numbers of injections are not statistically
different between aflibercept and ranibizumab they should be equalised.
Table 7 Previous aflibercept submissions: 1st year dosing assumptions
STA Date Aflibercept Ranibizumab
Wet AMD TA29427 May 2013 7.0 8.0
MO post CRVO TA30528 Dec 2013 8.3 8.8
DMO TA34629 May 2015 8.0 7.9
MO post BRVO* TA40930 Aug 2016 4.4 4.4
*Last 6 months of 1st year after laser
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
28
In previous STAs of aflibercept, the company has typically differentiated both the
numbers of injections in the 1st year and the clinical effectiveness estimates between
aflibercept and ranibizumab. Table 7 shows the base case dosing assumptions for
these previous aflibercept submissions and they are summarized here:
• For TA29427, the ERG subsequently equalised the numbers of injections.
• For TA30528, the company estimated 1st year means of 8.3 aflibercept
injections from GALILEO/COPERNICUS and 8.8 ranibizumab injections
from the published trial. The ERG only equalised these in a scenario analysis.
• For TA34629, the company estimated 1st year means of XXX aflibercept
injections from VIVID/VISTA and 7.40 ranibizumab injections from
RESTORE/REVEAL but applied 8.0 for aflibercept based upon the SmPC and
7.9 for ranibizumab based upon the mixed treatment comparison. The ERG
suggested estimates of XXX injections for aflibercept and 7.4 for ranibizumab.
• Only for TA40930 has the company equalised the 1st year injections between
aflibercept and ranibizumab, with ranibizumab being assumed to require the
same number of injections as aflibercept in the VIBRANT trial. The ERG
noted that the VIBRANT trial involved 9 aflibercept injections in the 1st year
for 1st line aflibercept compared to the BRAVO trial involving 8 ranibizumab
injections in the 1st year for 1st line ranibizumab, and undertook a scenario
analysis which reduced the number of 1st year 2nd line ranibizumab injections
by 1.
The FADs have, at times, suggested that there is little evidence of any real
difference in clinical effectiveness between aflibercept and ranibizumab.
The company surveyed 52 ophthalmologists all of whom treated a minimum of 2
mCNV cases annually. This survey estimated that there would be very much
fewer aflibercept injections than ranibizumab injections.
Table 8 Company resource use survey: injection frequency medians
Year 1 Year 2
Aflibercept XXX XXX
Ranibizumab XXX XXX
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
29
In the opinion of the ERG, given the trials’ injection frequencies, the estimates in
Table 8, while medians, are not credible.
In light of the company’s previous approaches, it may be questionable for it to now
switch to demanding a statistically significant difference in the numbers of injections
in the 1st year for them to be differentiated.
Table 9 Company trial injection estimates, standard deviations and 95% CIs
N Mean Inj. SD 95% CI
Aflibercept (AFLI) 91 4.2 XX (3.56, 4.84)
Ranibizumab visual acuity (RANI VA) 106 4.6 2.6 (4.10, 5.10)
Ranibizumab disease activity (RANI DA) 116 3.5 3.0 (2.95, 4.05)
Figure 3 1st year mean numbers of injections
The ERG’s clinical expert is of the opinion that the MYRROR26 retreatment criteria
are closer to the RADIANCE25 disease activity retreatment arm than the RADIANCE
visual acuity retreatment arm. Ranibizumab was also approved for mCNV based upon
modelling of the disease activity retreatment arm.
0
1
2
3
4
5
6
AFLI RANI DA RANI VA
Trials' mean number of injections: 1st year
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
30
ERG expert opinion is that the trials were relatively small, differed in terms of
populations and sites and also in the protocols for the decision to retreat, and that, as a
consequence, the number of aflibercept injections is unlikely to differ much from that
of ranibizumab.
3.3 Injection frequency and patient drop outs
The patient flow within MYRROR and RADIANCE is presented below.
Table 10 MYRROR and RADIANCE patient flow
MYRROR26 RADIANCE25
Aflibercept SHAM
Ranibizumab
visual acuity
Ranibizumab
disease activity vPDT
Baseline 91 31 106 116 55
Lack of efficacy .. .. 1 0 0
Adverse event X X .. .. ..
Patient withdrawal X X 1 2 0
Protocol violation X X 1 1 0
Other treatment 0 X .. .. ..
LTFU 0 0 3 1 0
EoT 78 (86%) 24 (77%) 100 (94%) 112 (97%) 55 (100%)
*Treatment failure
The 86% remaining in the MYRROR aflibercept arm at end of trial is reasonably
different from the 97% remaining in the RADIANCE disease activity arm, and
the difference is statistically significant (p=0.01).
Based upon figures 1 and 8 of the company submission to TA29816 and an
assumption that 1 month is 4 weeks, the ERG takes the period from day 1 until
prior to month 12 during the RADIANCE trial to span the same amount of time as
the MYRROR trial. Patient numbers and injections data supplied by the company
at clarification are shown in Table 11. Note that the percentages for aflibercept
injections are the number of injections divided by the baseline patient number and
not the contemporaneous patient numbers remaining within the trial. The patient
numbers and mean injections, as reported in Wolf et al25, are also given for
RADIANCE.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
31
The mean number of injections for aflibercept is calculated as the total number of
injections, 379, divided by the baseline number who received an injection, 90, to
yield an estimate of an average of 4.2 aflibercept injections. Other things being
equal, the total number of injections will fall as the number of patients remaining
within the trial falls.
Table 11 MYRROR and RADIANCE dosing
MYRROR26 RADIANCE25
Week Aflibercept Month Ranibizumab
visual acuity
Ranibizumab disease
activity
N Injections N N
0 90 90 (100%) 0 106* 116
4 XX XXX 1 .. ..
8 XX XXX 2 .. ..
12 XX XXX 3 105 116
16 XX XXX 4 .. ..
20 XX XXX 5 .. ..
24 83 XXX 6 103 116
28 XX XXX 7 .. ..
32 XX XXX 8 .. ..
36 XX XXX 9 .. ..
40 XX XXX 10 .. ..
44 XX XXX 11 .. ..
48 78 12 100 112
Mean inj 4.2 4.6 3.5
* 1 patient withdrew without any post baseline assessment so was excluded from
the efficacy calculations
Using a crude linear interpolation to infer the monthly RADIANCE patient
numbers, the average duration of follow-up is around 5% higher in the
ranibizumab visual acuity arm and around 7% higher in the ranibizumab disease
activity arm than that of the MYRROR aflibercept arm. This might argue for
increasing the mean number of aflibercept injections during the 1st year by a
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
32
similar amount, to between 4.4 and 4.5. None of this takes into account the
possible effects of patient attrition during the MYRROR and RADIANCE trials
upon clinical effectiveness estimates, and how the LOCF might be affected.
Wolf et al25 also noted that, during the last 6 months of RADIANCE, only 37% of
patients in the ranibizumab disease activity arm received any injections. Notably,
Wolf et al,25 which was supported by Novartis, do not supply the parallel 6 month
dosing figure for the ranibizumab visual acuity arm. During the last 6 months of
MYRROR, the percentage of patients receiving an injection each month was
typically around XXX. This may suggest lower dosing during the last 6 months of
the 1st year of treatment in the RADIANCE ranibizumab disease activity arm than
in the MYRROR aflibercept arm. If so, any lower dosing for ranibizumab might
flow into the 2nd year of treatment as well.
The ERG performed sensitivity analyses that apply the 1st year mean numbers of
injections that were observed in the trials, increase the 1st year mean numbers of
injections for aflibercept to 4.5 and a combination of these.
3.4 Administration costs
The ERG clinical expert has indicated that the ranibizumab prefilled syringe is more
convenient but, compared to using a pre-filled vial, has little practical impact upon
clinic time and the numbers of patients that can be treated. The additional costs of the
needle and syringe required for each aflibercept administration compared to the
prefilled ranibizumab syringe have typically not been included in previous
assessments but are likely to be relatively minor.
As a consequence, administrations costs will only differ between aflibercept and
ranibizumab if the numbers of administrations differ. The previous aflibercept
STAs have made the following assumptions about administration and monitoring
costs:
• TA294:27 Wet AMD: Based upon a balance between one-stop administration
and monitoring visits and two-stop administration and monitoring visits. This
appears to have been largely superseded by the subsequent aflibercept STAs.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
33
• TA305:28 MO from CRVO: Administration costs based upon 52% as
outpatient and 48% as day case to give an average of £233.24. Monitoring
costs based upon BZ23Z plus the cost of a consultant-led outpatient
appointment yielding a total of £197.00. A one-stop model is applied where
administration visits double as monitoring visits suggesting a net cost for
administration compared to monitoring of £36.24.
• TA346:29 DMO: Administration costs based upon RD40Z 20 minute
ultrasound cost of £54.54, which is additional to the £139.22 cost per
monitoring visit.
• TA409:30 MO from BRVO: Administration costs based upon RD407 20
minute ultrasound cost of £53.96, which is additional to the £150.07 cost per
monitoring visit.
In the light of these precedents and the number of injections for mCNV in the 1st year
probably being below the total number of patient visits, the ERG will apply a net
administration cost of the 2015-16 reference costs for RD40Z of £55.14. Within the
2015-16 reference costs for outpatient procedures, the ERG cannot find BZ23Z, the
closest alternative appearing to be BZ87A Minor vitreous procedures age 19+ at an
average cost of £102. This will be applied as a sensitivity analysis. Based upon ERG
clinical expert opinion, the ERG will assume that the total number of patient visits is
the same for each treatment.
3.5 Monitoring frequency
The aflibercept SmPC states:22
The recommended dose for Eylea is a single intravitreal injection of 2 mg
aflibercept equivalent to 50 microlitres. Additional doses may be
administered if visual and/or anatomic outcomes indicate that the disease
persists. Recurrences should be treated as a new manifestation of the
disease. The schedule for monitoring should be determined by the treating
physician. The interval between two doses should not be shorter than one
month.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
34
The ranibizumab SmPC states:14
Treatment is initiated with one injection per month until maximum visual
acuity is achieved and/or there are no signs of disease activity i.e. no change
in visual acuity and in other signs and symptoms of the disease under
continued treatment. In patients with wet AMD, DME and RVO, initially, three
or more consecutive, monthly injections may be needed. Thereafter,
monitoring and treatment intervals should be determined by the physician and
should be based on disease activity, as assessed by visual acuity and/or
anatomical parameters.
ERG clinical expert opinion indicates that, provided that there is broad clinical
equivalence, there would be no difference in monitoring frequency between
ranibizumab and aflibercept, with a probable average of around 6 visits during the
1st year of treatment.
3.6 Bilateralism and recurrence
Based upon a resource use survey of 52 ophthalmologists, the company models XXX
bilateral involvement at baseline and so requiring treatment in both eyes. The
company also estimates an annual recurrence rate of XXX from its resource use
survey. Recurrence is assumed to require another year of treatment.
3.7 Cost comparison results
The company base case using list prices for both aflibercept and ranibizumab are
shown in Table 12a. The per-patient based estimates are the discounted costs of
the initial treatments of the first and second eye plus the discounted costs of the
treatment of recurrence in the first and second eye.
a There is a minor error in the calculation of the discount factors, the company model applying (1-r)t
rather than 1/(1+r)t. This has been corrected in what follows.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
35
Table 12 Company model per patient: at ranibizumab and aflibercept list prices
Drug Admin Total
Ranibizumab £12,594 £1,260 £13,854
Aflibercept £18,651 £1,260 £19,911
Net £6,057 £0 £6,057
The quite substantial additional cost per patient of £6,057 from aflibercept is due
to its list price being £265 more than the list price of ranibizumab.
The ERG revised base case adds £1,260 administration costs to both arms but does
not affect the net amounts. The ERG sensitivity analyses are shown in table 13. The
population based estimate is based upon the NICE estimate of an annual 2,917
incident eyes that will receive treatment which, when coupled with the company
estimate of 12.5% bilateralism at baseline, suggests an annual number of new patients
of 2,593. A 50% market share for aflibercept and a 50% market share for ranibizumab
are also assumed. This may be less realistic for the scenarios where one drug requires
more injections during the 1st year of treatment and recurrence than the other.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
36
Table 13 ERG sensitivity analyses at ranibizumab and aflibercept list prices
Drug Admin Total Population
Base case £6,057 £0 £6,057 £7.9mn
SA01: ranibizumab visual acuity dosing £4,914 -£114 £4,800 £6.2mn
SA02: ranibizumab disease activity dosing £8,057 £200 £8,257 £10.7mn
SA03: aflibercept 4.5 1st yr £7,326 £86 £7,412 £9.6mn
SA04: SA01 + £102 admin £4,914 -£212 £4,702 £6.1mn
SA05: SA02 + £102 admin £8,057 £370 £8,427 £10.9mn
SA06: SA03 + £102 admin £7,326 £159 £7,485 £9.7mn
SA07: No recurrence £1,531 £0 £1,531 £2mn
SA08: No bilateral £5,384 £0 £5,384 £7.9mn
SA09: SA07 + SA08 £1,361 £0 £1,361 £2.0mn
SA10: SA01 + SA09 £1,142 -£22 £1,120 £1.6mn
SA11: SA02 + SA09 £1,745 £38 £1,784 £2.6mn
SA12: SA03 + SA09 £1,605 £16 £1,621 £2.4mn
SA13: SA04 + SA09 £1,142 -£41 £1,101 £1.6mn
SA14: SA05 + SA09 £1,745 £71 £1,816 £2.6mn
SA15: SA06 + SA09 £1,605 £30 £1,635 £2.4mn
SA16: SA02 + SA03 £9,326 £286 £9,612 £12.5mn
SA17: SA16+ SA09 £1,989 £55 £2,044 £3.0mn
SA18: SA16 + SA09 + £102 admin £1,989 £102 £2,090 £3.0mn
SA19: SA16 + SA07 £2,237 £62 £2,299 £3.0mn
As would be expected if the ranibizumab visual acuity dosing is applied [SA01],
this increases costs in the ranibizumab arm and aflibercept results in a slightly
smaller cost of £4,800 per patient. The annual incident population present value
of the costs is estimated to be £6.2mn. However, if the ranibizumab disease
activity dosing [SA02] is applied, aflibercept is still more costly by £8,257 per
patient, which translates into an annual incident population present value cost of
£10.7mn.
If aflibercept dosing in the 1st year should be adjusted for the statistically higher
drop-out rate during MYRROR to 4.5 injections [SA03], it is more costly than
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
37
ranibizumab by £7,412 per patient. This still assumes the 4.2 1st year injections
for ranibizumab. If these are reduced to the 3.5 of the ranibizumab disease activity
arm [SA16], the increase in costs associated with aflibercept rises to £9,612 per
patient and the annual incident population present value cost to £12.5mn.
Applying the 1st year injections frequencies of 4.5 for aflibercept and 3.5 for
ranibizumab [SA16-SA19] can be seen as the worst case 1st year injections
frequencies scenario for aflibercept given the trials’ data. Differentiation by the
number of 1st year injections might also argue for differentiating the numbers of
subsequent injections.
If the 1st year numbers of injections are differentiated between aflibercept and
ranibizumab, there are reasonable differences in total costs. The XXX recurrence
and retreatment rate estimate is a key determinant of the total lifetime costs, and,
by implication, the net lifetime patient costs if treatments are differentiated by the
number of 1st year injections. The worst case 1st year injections frequencies
scenario for aflibercept [SA16] estimated net cost of £9,612 falls to only £2,299
[SA19] if the rate of recurrence is set to 0%.
ERG clinical expert opinion views the XXX recurrence and retreatment as
reasonable, certainly in the early years. The implied total number of anti-VEGF
treatments over the patient lifetime exceeds that modelled by the company for
other conditions: an initial 5.2 injections per eye followed by a further 24.6
injections for treatment of recurrenceb, suggesting a total lifetime number of
injections of 29.7 per eye. The company has ignored the 1st and 2nd year injection
frequency estimates of its expert survey.
Given the XXX recurrence and retreatment assumptions, the model will also
show some sensitivity to both the baseline age and the proportion of female
patients since this will determine survival which rolls through to the costs of
recurrence. Given that the MYRROR trial was in an Asian population, there
might be some questions about its generalisability, but the baseline age and
b These injections may also to some degree cover further bilateral incidence further to the baseline XXXX prevalence estimate.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
38
proportion of women of 58 years and 76%, respectively, are similar to the 56
years and 75% of RADIANCE.
3.8 Differentiating aflibercept from ranibizumab in a cost utility analysis
This section is highly speculative. It is intended to highlight the possible
approaches to differentiating aflibercept from ranibizumab through the cost utility
modelling that would be required for an STA, and the approaches the company
might explore. It does not provide formal estimates of the cost effectiveness of
aflibercept compared to ranibizumab.
The costing considerations outlined above do not address any effects upon patient
gains. The company indirect treatment comparison suggests that aflibercept is
better, though not statistically significantly better, than both ranibizumab visual
acuity retreatment and ranibizumab disease activity retreatment.
Table 14 Company indirect treatment comparison - mean letters gain in BCVA
treated eye at 3 months
Mean SD 95% CI
Aflibercept vs ranibizumab visual acuity 1.34 3.40 (-5.35, 8.00)
Aflibercept vs ranibizumab disease activity 0.94 3.38 (-5.67, 7.56)
Due to the VIP trial not reporting the proportions gaining and losing 10 letters and
15 letters it is not possible to undertake an indirect treatment comparison for these
variables. Any modelling of the cost effectiveness of aflibercept versus
ranibizumab that adopted the modelling framework of the ranibizumab
assessment16 would have to infer the relative risks of gaining and losing letters
from the mean letters change and some mapping from this to the proportions of
patients gaining and losing letters. This would introduce a reasonable amount of
uncertainty to any cost effectiveness modelling that adopted the modelling
framework of the ranibizumab assessment.16
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
39
The alternative would seem to be to develop a new model based upon the mean
letters change in BCVA at 3 months. It seems likely that this would immediately
suggest that aflibercept dominates the ranibizumab visual acuity dosing arm.
However, it should be borne in mind that ranibizumab was approved for mCNV
based upon modelling of the ranibizumab disease activity dosing arm.
Previous assessments have often used the Czoski-Murray mapping from changes
in the patients’ overall BCVA to quality of life.31 The mean gain of 0.94 letters
for aflibercept versus the ranibizumab disease activity arm would translate to a
quality of life gain of 0.007. The gains in visual acuity with the anti-VEGFs are
relatively rapid and then tend to plateau and, assuming that the gain at three
months would apply for the remainder of the year, may be reasonable. The XXX
recurrence and retreatment rate might justify an assumption of the gain being
maintained for the remainder of the patient lifetime. While this might be
optimistic, it appears that it would suggest a total patient gain of 0.114 QALYs
from aflibercept compared to ranibizumab if the gain was in the patient’s bilateral
BCVA. A willingness to pay of £30k/QALY would justify an additional expense
of £3,420. If the gain was in the worse seeing eye, previous assessments have
suggested it might be only 30% of this, and so only warrant an additional expense
of £1,026. These amounts can be compared with the estimates of the net
additional costs and savings per patient associated with aflibercept under the
various scenarios outlined above.
Aflibercept would dominate ranibizumab visual acuity retreatment at central
estimates. Whether aflibercept is cost effective compared to ranibizumab disease
activity retreatment at central estimates is less clear. It may depend upon the
proportion that has their better seeing eye treated, and so upon the baseline XXX
bilateralism rate and whether further bilateralism develops post baseline.
Any move to an STA and the associated cost-utility modelling could result in the
company revising or limiting the duration of its XXX recurrence rate estimate. ERG
clinical expert opinion suggests that the XXX estimate is reasonable, at least in the
short term. Limiting its duration might have a limited impact upon cost effectiveness
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
40
estimates since there would probably have to be a parallel limitation on patient
benefits.
Any move to an STA could result in the company modelling an incidence of
bilateralism post baseline. This would increase the proportion of patients being treated
in their better seeing eye. Net costs would increase, unless there were significant cost
offsets from reduced blindness which seems unlikely to the ERG. But the net benefits
would increase more due to the better seeing eye being treated and the cost
effectiveness of aflibercept would improve.
Given the uncertainty about the clinical effectiveness estimates, the uncertainty about
the mean numbers of injections in the 1st year, the overlapping confidence intervals
and that indirect treatment comparison estimates of the proportions gaining and losing
letters are not possible any cost utility modelling would have a high degree of
uncertainty around it.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
41
4 ERG’s overall view of the company’s cost comparison case
The ERG opinion is that aflibercept is likely to offer similar benefits to ranibizumab
with a similar safety profile. A cost comparison was thus possible. The cost
comparison findings are sensitive to the number of injections, recurrence and
retreatment rates.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
42
5 References
1. Penn JS, Madan A, Caldwell RB, Bartoli M, Caldwell RW, Hartnett ME. Vascular endothelial growth factor in eye disease. Prog Retin Eye Res 2008;27:331-71. 2. Senger DR, Galli SJ, Dvorak AM, Perruzzi CA, Harvey VS, Dvorak HF. Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid. Science 1983;219:983-5. 3. Senger DR, Perruzzi CA, Feder J, Dvorak HF. A highly conserved vascular permeability factor secreted by a variety of human and rodent tumor cell lines. Cancer Res 1986;46:5629-32. 4. Neelam K, Cheung CM, Ohno-Matsui K, Lai TY, Wong TY. Choroidal neovascularization in pathological myopia. Prog Retin Eye Res 2012;31:495-525. 5. Ng DS, Kwok AK, Chan CW. Anti-vascular endothelial growth factor for myopic choroidal neovascularization. Clin Experiment Ophthalmol 2012;40:e98-e110. 6. Wang E, Chen Y. Intravitreal anti-vascular endothelial growth factor for choroidal neovascularization secondary to pathologic myopia: systematic review and meta-analysis. Retina 2012;33:1375-92. 7. Chan WM, Lai TY, Chan KP, Li H, Liu DT, Lam DS, et al. Changes in aqueous vascular endothelial growth factor and pigment epithelial-derived factor levels following intravitreal bevacizumab injections for choroidal neovascularization secondary to age-related macular degeneration or pathologic myopia. Retina 2008;28:1308-13. 8. Bandello F, Casalino G, Loewenstein A, Goldstein M, Pelayes D, Battaglia Parodi M. Pharmacological approach to diabetic macular edema. Ophthalmic Res 2014;51:88-95. 9. Holash J, Davis S, Papadopoulos N, Croll SD, Ho L, Russell M, et al. VEGF-Trap: a VEGF blocker with potent antitumor effects. Proc Natl Acad Sci U S A 2002;99:11393-8. 10. Ohr M, Kaiser PK. Intravitreal aflibercept injection for neovascular (wet) age-related macular degeneration. Expert Opin Pharmacother 2012;13:585-91. 11. Papadopoulos N, Martin J, Ruan Q, Rafique A, Rosconi MP, Shi E, et al. Binding and neutralization of vascular endothelial growth factor (VEGF) and related ligands by VEGF Trap, ranibizumab and bevacizumab. Angiogenesis 2012;15:171-85. 12. Sharma YR, Tripathy K, Venkatesh P, Gogia V. Aflibercept - how does it compare with other anti-VEGF drugs? Austin J Clin Ophthalmol 2014;1:1016.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
43
13. Chappelow AV, Kaiser PK. Neovascular age-related macular degeneration: potential therapies. Drugs 2008;68:1029-36. 14. Lucentis 10 mg/ml solution for injection : Summary of product characteristics [document on the Internet]: Electronic Medicines Compendium; 2016. http://www.medicines.org.uk/emc/medicine/19409#NUMBER [Accessed July 2017] 15. Ferrara N, Damico L, Shams N, Lowman H, Kim R. Development of ranibizumab, an anti-vascular endothelial growth factor antigen binding fragment, as therapy for neovascular age-related macular degeneration. Retina 2006;26:859-70. 16. Ranibizumab for treating choroidal neovascularisation associated with pathological myopia. NICE Technology Appraisal Guidance TA298 London: National Institute for Health and Care Excellence; 2013. https://www.nice.org.uk/guidance/ta298 [Accessed July 2017] 17. Schmidt-Erfurth U, Laqua H, Schlotzer-Schrehard U, Viestenz A, Naumann GO. Histopathological changes following photodynamic therapy in human eyes. Arch Ophthalmol 2002;120:835-44. 18. Blinder KJ, Blumenkranz MS, Bressler NM, Bressler SB, Donato G, Lewis H, et al. Verteporfin therapy of subfoveal choroidal neovascularization in pathologic myopia: 2-year results of a randomized clinical trial--VIP report no. 3. Ophthalmology 2003;110:667-73. 19. Verteporfin in Photodynamic Therapy Study G. Photodynamic therapy of subfoveal choroidal neovascularization in pathologic myopia with verteporfin. 1-year results of a randomized clinical trial--VIP report no. 1. Ophthalmology 2001;108:841-52. 20. Giansanti F, Virgili G, Donati MC, Giuntoli M, Pieretti G, Abbruzzese G, et al. Long-term results of photodynamic therapy for subfoveal choroidal neovascularization with pathologic myopia. Retina 2012;32:1547-52. 21. Hayashi K, Ohno-Matsui K, Shimada N, Moriyama M, Hayashi W, Wang J, et al. Long-term results of photodynamic therapy for choroidal neovascularization in Japanese patients with pathologic myopia. Am J Ophthalmol 2011;151:137-47.e1. 22. Eylea 40mg/ml solution for injection in a vial: summary of product characteristics: Electronic Mediicnes Compendium (eMC); 2016. http://www.medicines.org.uk/emc/medicine/27224 [Accessed July 2017] 23. Wong TY, Ferreira A, Hughes R, Carter G, Mitchell P. Epidemiology and disease burden of pathologic myopia and myopic choroidal neovascularization: An evidence-based systematic review. Am J Ophthalmol 2014;157:9-25.e12. 24. Wong TY, Ohno-Matsui K, Leveziel N, Holz FG, Lai TY, Yu HG, et al. Myopic choroidal neovascularisation: current concepts and update on clinical management. Br J Ophthalmol 2015;99:289-96.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
CONFIDENTIAL UNTIL PUBLISHED
44
25. Wolf S, Balciuniene VJ, Laganovska G, Menchini U, Ohno-Matsui K, Sharma T, et al. RADIANCE: a randomized controlled study of ranibizumab in patients with choroidal neovascularization secondary to pathologic myopia. Ophthalmology 2014;121:682-92.e2. 26. Ikuno Y, Ohno-Matsui K, Wong TY, Korobelnik JF, Vitti R, Li T, et al. Intravitreal Aflibercept Injection in Patients with Myopic Choroidal Neovascularization: The MYRROR Study. Ophthalmology 2015;122:1220-7. 27. Aflibercept solution for injection for treating wet age‑related macular degeneration. NICE Technology Appraisal Guidance TA294. . London: National Institute for Health and Care Excellence; 2013. https://www.nice.org.uk/guidance/ta294 [Accessed July 2017] 28. Aflibercept for treating visual impairment caused by macular oedema secondary to central retinal vein occlusion. NICE Technology Appraisal Guidance TA305. London: National Institute for Health and Care Excellence; 2014. https://www.nice.org.uk/guidance/ta305 [Accessed July 2017] 29. Aflibercept for treating diabetic macular oedema. NICE Technology Appraisal Guidance TA 346. London: National Institute for Health and Care Excellence; 2015. URL: https://www.nice.org.uk/guidance/ta346 [Accessed July 2017] 30. Aflibercept for treating visual impairment caused by macular oedema after branch retinal vein occlusion. NICE Technology Appraisal Guidance TA409. London: National Institute for Health and Care Excellence; 2016. https://www.nice.org.uk/guidance/ta409 [Accessed July 2017] 31. Czoski-Murray C, Carlton J, Brazier J, Young T, Papo NL, Kang HK. Valuing condition-specific health states using simulation contact lenses. Value Health 2009;12:793-9.
Copyright 2017 Queen's Printer and Controller of HMSO. All rights reserved.
Top Related